Earlier quoted context omitted.
Not a go fan, but seems like if anyone has scale problems, google would, and go would give them these issues too which is surprising since that was kind of the goal to solve from their perspective.
There's different definitions of scale. He is talking about from an engineering perspective and at Google we have seen nothing other than a few periphery sites being written in Go. I fully agree that Go is a poor language if you are building a large, monolithic application.
I find that statement hilarious considering that it was explicitly one of the design considerations[1] for Go. So they either failed massively, or there's a difference in perspective on what a good language for programming in the large is.
From Rob Pike's talk[1]:
> Go was designed to address the problems faced in software development at Google, which led to a language that is not a breakthrough research language but is nonetheless an excellent tool for engineering large software projects.